Shot Ingestion Flexibility

[tasansal]

Intro

Raw seismic exploration datasets are called shots, and they're typically stored in SEG-Y files.

The below concept represents the geometry of a 3D shot with a lot fewer streamers (aka. cables) and receivers (aka. channels) than an actual survey:

Cable 1 ->        / 1------------------20
Cable 2 ->       / 1------------------20
.               / 1------------------20
.          ⛴ ☆ - 1------------------20
.               \ 1------------------20
Cable 6 ->       \ 1------------------20
Cable 7 ->        \ 1------------------20

The ☆ represent an airgun array (aka. source or shot), and the - represents a channel on a streamer. In this example, we have a shot with 7 streamers, each with 20 receivers.

Problem

SEG-Y Revision 2.0 came out in 2017 with standards describing 3D offshore seismic surveys. However, any SEG-Y file made before that has no standard, and vendors may not provide all the necessary information in the headers.

We typically see 3D offshore shot datasets with the following pathologies:

Type	Streamer ID	Receiver ID	Wrapped
A	✓	✓	✓
B	✓	✓	✗
C	✗	✓	✗

Type A

This is the case that is showcased at the beginning. The SEG-Y file has the following

Cable:   1 1 ...  1  1 | 2 2 ...  2  2 | 3 3 ...  3  3 | ...... | ...  7  7 
Channel: 1 2 ... 19 20 | 1 2 ... 19 20 | 1 2 ... 19 20 | ...... | ... 19 20

Type B trace headers show:

Cable 1 ->        / 1------------------20
Cable 2 ->       / 21-----------------40
.               / 41-----------------60
.          ⛴ ☆ - 61-----------------80
.               \ 81----------------100
Cable 6 ->       \ 101---------------120
Cable 7 ->        \ 121---------------140

The SEG-Y file trace headers are like below:

Cable:   1 1 ...  1  1 |  2  2 ...  2  2 |  3  3 ...  3  3 | ...... | ...   7   7 
Channel: 1 2 ... 19 20 | 21 22 ... 39 40 | 41 42 ... 59 60 | ...... | ... 139 140

The translation here is simple: chan = (chan - 1) % N + 1 where N is the "channels per cable" which must be known.

Type C trace headers show:

Cable ? ->        / 1------------------20
Cable ? ->       / 21-----------------40
.               / 41-----------------60
.          ⛴ ☆ - 61-----------------80
.               \ 81----------------100
Cable ? ->       \ 101---------------120
Cable ? ->        \ 121---------------140

Cable translation: cable = floor((chan - 1) / N) + 1 where N is the "channels per cable" which must be known.
and we do the channels after this with the same formula as Type B.

Solution

To be able to ingest these legacy files to the correct geometry, we need to do some grid
manipulations. This PR addresses that.

We can specify options to:

• Calculate the streamer numbers (from the original unwrapped channel and by knowing receivers per streamer)
• Calculate the wrapped receiver numbers (from the original unwrapped channel and by knowing receivers per streamer)

Example (proposed) usages

Type A

Simplest, natural geometry of the file with cable numbers and wrapped channels.

segy_to_mdio(
    segy_path="prefix/shot_file.segy",
    mdio_path_or_buffer="s3://bucket/shot_file.mdio",
    index_bytes=(17, 137, 13),
    index_lengths=(4, 2, 4),
    index_names=("shot", "cable", "channel"),
    chunksize=(8, 2, 128, 1024),
)

Type B

Cable numbers exist, but channels numbers are unwrapped.
Note the addition of grid_overrides and the keywords ChannelWrap and ChannelsPerCable.

segy_to_mdio(
    segy_path="prefix/shot_file.segy",
    mdio_path_or_buffer="s3://bucket/shot_file.mdio",
    index_bytes=(17, 137, 13),
    index_lengths=(4, 2, 4),
    index_names=("shot", "cable", "channel"),
    chunksize=(8, 2, 128, 1024),
    grid_overrides={"ChannelWrap": True, "ChannelsPerCable": 800},
)

Type C

No cable numbers were provided, and unwrapped channels.
Note the modification of grid_overrides with an additional keyword: CalculateCable.
Also note that the "channel" header index bytes and lengths are set to None.

segy_to_mdio(
    segy_path="prefix/shot_file.segy",
    mdio_path_or_buffer="s3://bucket/shot_file.mdio",
    index_bytes=(17, None, 13),
    index_lengths=(4, None, 4),
    index_names=("shot", "cable", "channel"),
    chunksize=(8, 2, 128, 1024),
    grid_overrides={"ChannelWrap": True, "ChannelsPerCable": 800, "CalculateCable": True},
)

[codecov-commenter]

Codecov Report

Merging #180 (0824783) into main (c507dee) will decrease coverage by 1.02%.
The diff coverage is 4.34%.

📣 This organization is not using Codecov’s GitHub App Integration. We recommend you install it so Codecov can continue to function properly for your repositories. Learn more

@@            Coverage Diff             @@
##             main     #180      +/-   ##
==========================================
- Coverage   81.62%   80.60%   -1.02%     
==========================================
  Files          43       43              
  Lines        1654     1676      +22     
  Branches      179      189      +10     
==========================================
+ Hits         1350     1351       +1     
- Misses        267      285      +18     
- Partials       37       40       +3     

Impacted Files	Coverage Δ
src/mdio/converters/segy.py	75.36% <ø> (ø)
src/mdio/segy/_workers.py	20.89% <0.00%> (-2.06%)	⬇️
src/mdio/segy/utilities.py	60.78% <6.25%> (-24.93%)	⬇️

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.